Compositional data analysis refers to analyzing relative information, based on ratios between the variables in a data set. Data from epidemiology are usually treated as absolute information in an analysis. We outline the differences in both approaches for univariate and multivariate statistical analyses, using illustrative data sets from Austrian districts. Not only the results of the analyses can differ, but in particular the interpretation differs. It is demonstrated that the compositional data analysis approach leads to new and interesting insights.
LincolnTLBuilderC. Global healthcare and the flux of technology. Int J Med Informat1999; 53: 213224–213224.
2.
HennekensCBuringJ. Epidemiology in medicine, Philadelphia, PA: Lippincott, Williams & Wilkins, 1987.
3.
PearsonK. Mathematical contributions to the theory of evolution—On a form of spurious correlation which may arise when indices are used in the measurement of organs. Proceedings of the Royal Society of London1897; LX: 489–502.
4.
AitchisonJ. The statistical analysis of compositional data, London: Chapman & Hall, 1986.
5.
Pawlowsky-GlahnVEgozcueJJTolosana-DelgadoR. Modeling and analysis of compositional data, Chichester: Wiley, 2015.
6.
LeiteMC. Applying compositional data methodology to nutritional epidemiology. Stat Methods Med Res2014; 24: 43–70.
7.
TsilimigrasMCFodorAA. Compositional data analysis of the microbiome: fundamentals, tools, and challenges. Ann Epidemiol2016; 26: 330–335.
8.
MaronnaRMartinDYohaiV. Robust statistics: theory and methods, Toronto, ON: John Wiley & Sons Canada Ltd., 2006.
9.
BeaumontJLixLYostK, et al.Application of robust statistical methods for sensitivity analysis of health-related quality of life outcomes. Qual Life Res2006; 15: 349–356.
10.
EgozcueJJPawlowsky-GlahnVMateu-FiguerasG, et al.Isometric logratio transformations for compositional data analysis. Math Geol2003; 35: 279–300.
11.
FišerováEHronK. On the interpretation of orthonormal coordinates for compositional data. Math Geosci2011; 43: 455–468.
12.
EgozcueJJPawlowsky-GlahnV. Groups of parts and their balances in compositional data analysis. Math Geol2005; 37: 795–820.
13.
Pawlowsky-GlahnVBucciantiA. Compositional data analysis: theory and applications, Chichester: Wiley, 2011.
14.
JolliffeI. Principal component analysis, London: John Wiley & Sons, 2014.
15.
GabrielKR. The biplot graphic display of matrices with application to principal component analysis. Biometrika1971; 58: 453–467.
16.
FilzmoserPHronKReimannC. Principal component analysis for compositional data with outliers. Environmetrics2009; 20: 621–632.
17.
AitchisonJGreenacreM. Biplots of compositional data. J R Stat Soc Ser C (Appl Stat)2002; 51: 375–392.
18.
KynčlováPFilzmoserPHronK. Compositional biplots including external non-compositional variables. Statistics2016; 50: 1132–1148.
19.
ReimannCFilzmoserPGarrettR, et al.Statistical data analysis explained: applied environmental statistics with R, Chichester: Wiley, 2008.
20.
MahalanobisP. On the generalised distance in statistics. Proc Natl Inst Sci India1936; A2: 49–55.
21.
YohaiVJ. High breakdown-point and high efficiency robust estimates for regression. Ann Stat1987; 15: 642–656.
22.
RousseeuwPJvan ZomerenBC. Unmasking multivariate outliers and leverage points. J Am Stat Assoc1990; 85: 633–639.
23.
FilzmoserPHronK. Outlier detection for compositional data using robust methods. Math Geosci2008; 40: 233–248.
